EP3145140A1 - Nonlinear system distortion correction device and method - Google Patents

Abstract

The present invention discloses an apparatus and method for nonlinear system distortion correction. The apparatus includes: apparatus for nonlinear system distortion correction including: a robustness signal source generator, arranged to generate a robustness signal provided to an adaptive device; the adaptive device, arranged to collect a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link, and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and a pre-corrector, arranged to according to the correction parameter perform pre-correction processing on the main link signal output by the signal generator. The present invention ensures, by providing the robustness signal, that, a system will not bring the problem of unknowable abrupt change in power or abnormal output frequency spectrum, which is caused by the fact that the difference between a signal characteristic in a former configuration and a signal characteristic in a latter configuration is too large under various dynamic configurations after an actual link signal passes through a pre-distortion module, thereby improving the robustness of the system in the case of dynamic operation, and being able to satisfy various frequency band configuration requirements of operators and dynamic change requirements of telephone traffic in different regions.

Description

Technical Field
• The present invention relates to the field of communication technology, and in particular, to an apparatus and method for nonlinear system distortion correction.
Background
• With the development of mobile communications and increasingly scarce spectrum resources, it is often difficult for operators to obtain frequency spectrums with enough bandwidths within one frequency band. Each operator will more frequently establish a cross-band communication system in the future. This will cause communication device systems which are required to process cross-band ultra-wideband signals. Meanwhile, the operators' requirements for capital expenditure (CAPEX) and operating expense (OPEX) will gradually become a primary consideration when devices are selected. For a wireless communication base station system, about 80% of power consumption is generated by a radio frequency power amplifier (PA) in a radio remote unit (RRU) module. Thus, with the development of the digital mobile communication technology, high-efficiency power amplifiers (hereinafter referred to as power amplifiers) become a requirement which each of major system device manufacturers must meets. A major problem which the high-efficiency power amplifiers face is intermodulation interference generated when the power amplifiers operating close to a saturation region in a modem high-efficiency modulation mode, resulting in the power amplifiers generating serious nonlinear distortion. The digital pre-distortion technology becomes a basic power amplifier linearization function module of RRU systems in the current mainstream communication devices. In summary, the digital pre-distortion technology in the RRU systems is required to cope with cross-band ultra-wideband signals under the current operating condition.
• This condition brings a challenge to systems using the power amplifier linearization technology. All the current digital pre-distortion technologies are based on an off-line self-adaptive and iterative technique in which digital pre-distortion information cannot always completely follows the change in actual links. This problem appears to be particularly prominent under the configuration of cross-band ultra-wideband signals. For instance, it may be discovered in the actual test that, in the mode mixing cross-band application of Time Division-Synchronous Code Division Multiple Access (TD-SCDMA) signals and Time Division Duplexing-Time Division Long Term Evolution (TDD-LTE) signals, under certain configuration conditions, when a cell is initially established, data which is set by a predistorter to extract linearization parameters is single frequency band because of no service in service timeslots of a certain frequency band. When services in the frequency band are increased, the predistorter cannot update the linearization parameters in real time, after the cross-band signals in the actual test go through the linearization parameters extracted by the single frequency band, an abrupt change of more than ten dBs will occur in output power, and the frequency spectrum is abnormal, which easily causes damage to the power amplifiers of the systems and huge harm to system robustness.
Content of the Invention
• Embodiments of the present invention provide an apparatus and method for nonlinear system distortion correction to at least solve the problem of poor system robustness in the prior art.
• In order to solve the above technical problem, on the one hand, the present invention provides an apparatus for nonlinear system distortion correction, including:
• a robustness signal source generator, arranged to generate a robustness signal provided to an adaptive device;
• the adaptive device, arranged to collect a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link, and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and
• a pre-corrector, arranged to according to the correction parameter perform pre-correction processing on the main link signal output by the signal generator.
• Further, the adaptive device includes:
• a data collection unit, arranged to collect the main link signal output by the signal generator and the feedback signal returned by the main link signal via the feedback link;
• a matrix composition unit, arranged to generate a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generate a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal, and perform matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z; and
• a correction parameter generation unit, arranged to extract a linearization parameter according to the matrixes R and Z, and use the linearization parameter as the correction parameter to be loaded into the pre-corrector.
• Further, the pre-corrector includes:
• an index address acquisition unit, arranged to calculate an amplitude value or power of the main link signal to obtain an index address of a correction signal;
• a correction signal acquisition unit, arranged to look up a corresponding lookup table according to the index address, and generate the correction signal according to contents of the lookup table; and
• a pre-correction processing unit, arranged to perform the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter.
• Further, the pre-correction processing unit performs the pre-correction processing according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$

  

  $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$

  

  $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$

  

  where, y is a pre-corrected signal; n is a signal sampling time serial number; F_{U, X}(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_{U, X}(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
• Further, the matrix composition unit performs the matrix composition using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

  

  $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

  

  where, a and b are weighting coefficients.
• Further, the robustness signal source generator is arranged to generate a robustness signal by the following way:
• performing frequency conversion, filtering and frequency shift processing on a baseband signal to obtain the main link signal, and performing data property processing on the main link signal to obtain the robustness signal.
• Further, the adaptive device performs the data property processing on the main link signal and the feedback signal after collecting the main link signal and the feedback signal.
• Further, the data property processing refers to performing operations of delay aligning and energy aligning on the collected signals.
• Further, the apparatus further includes:
• a digital to analog converter, arranged to convert a digital signal output by the pre-corrector into an analog signal;
• a nonlinear system, arranged to perform nonlinear processing on the analog signal; and an analog to digital converter, arranged to couple a signal out from the output signal by the nonlinear system, and convert the signal into a digital signal which is used as the feedback signal to be input into the adaptive device.
• On the other hand, the present invention further provides a method for nonlinear system distortion correction, including:
• generating a robustness signal provided to an adaptive device;
• collecting a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link, and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and
• performing, according to the correction parameter, pre-correction processing on the main link signal output by the signal generator.
• Further, generating a correction parameter further includes:
• collecting the main link signal output by the signal generator and the feedback signal returned by the main link signal via the feedback link;
• generating a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generating a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal, and performing matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z; and
• extracting a linearization parameter according to the matrixes R and Z, and using the linearization parameter as the correction parameter to be loaded into a pre-corrector.
• Further, the pre-correction processing further includes:
• calculating an amplitude value or power of the main link signal to obtain an index address of a correction signal;
• looking up a corresponding lookup table according to the index address, and generating the correction signal according to contents of the lookup table; and
• performing the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter.
• Further, the pre-correction processing is performed according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$

  

  $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$

  

  $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$

  

  where, y is a pre-corrected signal; n is a signal sampling time serial number; F_{U, X}(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_{U, X}(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
• Further, the matrix composition is performed using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

  

  $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

  

  where, a and b are weighting coefficients.
• Further, generating a robustness signal includes:
• performing frequency conversion, filtering and frequency shift processing on a baseband signal to obtain the main link signal and performing data property processing on the main link signal to obtain the robustness signal.
• Further, after collecting the main link signal and the feedback signal, the method further includes:
• performing the data property processing on the main link signal and the feedback signal.
• Further, the data property processing refers to performing operations of delay aligning and energy aligning on the collected signals.
• Further, a digital signal after the pre-correction processing is converted into an analog signal;
  nonlinear processing is performed on the analog signal; and
  a signal is coupled out from the signal after the nonlinear processing and is converted into a digital signal to be returned as the feedback signal.
• The beneficial effects of the present invention are as follows:
• The present invention ensures, by providing a robustness signal, that a system will not bring the problem of unknowable abrupt change in power or abnormal output frequency spectrum, which is caused by the fact that the difference between a signal characteristic in a former configuration and a signal characteristic in a latter configuration is too large under various dynamic configurations after an actual link signal passes through a pre-distortion module, thereby greatly improving the robustness of the system in the case of dynamic operation, and being able to satisfy various frequency band configuration requirements of operators and dynamic change requirements of telephone traffic in different regions.
Brief Description of Drawings
• FIG. 1 is a schematic diagram of a structure of an apparatus for nonlinear system distortion correction in an embodiment of the present invention;
• FIG. 2 is a schematic diagram of a structure of another apparatus for nonlinear system distortion correction in an embodiment of the present invention;
• FIG. 3 is a schematic diagram of a structure of an adaptive device in an embodiment of the present invention;
• FIG. 4 is a schematic diagram of a structure of a pre-corrector in an embodiment of the present invention;
• FIG. 5 is a schematic diagram of a structure of a further apparatus for nonlinear system distortion correction in an embodiment of the present invention; and
• FIG. 6 is a flow chart of a method for nonlinear system distortion correction in an embodiment of the present invention.
Preferred Embodiments of the Present Invention
• The present invention will be further described in detail in combination with the accompanying drawings and embodiments below. It should be understood that the specific embodiments described herein are only used to explain the present invention, not limit the present invention.
• The present invention puts forward a universal apparatus and method for improving system robustness using a power amplifier linearization technology, so as to, by using a method of adding a robustness signal source into a conventional pre-distortion self-iteration system to participate in extracting a linearization coefficient, solve the problem of damage to a power amplifier and abnormal signal frequency spectrum which may be caused due to unstable system output power caused by a pre-distortion module under some configurations in a cross-band ultra-wideband pre-distortion system.
• As shown in FIG. 1, an embodiment of the present invention relates to an apparatus for nonlinear system distortion correction including:
• a robustness signal source, generator arranged to generate a robustness signal provided to an adaptive device;
• the adaptive device, arranged to collect a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and
• a pre-corrector, arranged to, according to the correction parameter, perform pre-correction processing on the main link signal output by the signal generator.
• Further, as shown in FIG. 2, based on the above embodiment, the apparatus for nonlinear system distortion correction involved in the embodiment of the present invention can further include:
• a digital to analog converter (DAC) arranged to convert a digital signal output by the pre-corrector into an analog signal;
• a nonlinear system arranged to perform nonlinear processing on the analog signal; and an analog to digital converter (ADC) arranged to couple a signal out from the output signal by the nonlinear system, and convert the signal into a digital signal which is used as the feedback signal to be input into the adaptive device.
• The robustness signal source generator performs up-conversion and filtering on a baseband signal, then performs frequency shift processing according to a configuration required by the system, obtains a main link signal consistent with the bandwidth and configuration required by the system, collects the main link signal and performs data property processing, where the data property processing process is consistent with the data property processing in the adaptive device. The processed signal is stored and used as a robustness signal source to be supplied to the adaptive device for calling.
• As shown in FIG. 3, the adaptive device includes:
• a data collection unit arranged to collect the main link signal (i.e., a digital signal in a forward link) and the feedback signal (i.e., a digital signal in a power amplifier feedback link) returned by the main link signal via the feedback link;
• a signal processing unit arranged to perform data property processing on the collected main link signal and feedback signal, the data property processing referring to performing operations of delay aligning and energy aligning on the collected signals; and
• a matrix composition unit arranged to generate a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generate a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal and perform matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z.
• The matrix composition unit establishes a distortion model of the nonlinear system, and the distortion model uses a universal memory multinomial model, as shown in the following formula: $${\mathrm{S}}_{\mathrm{o}}\left(\mathrm{n}\right)=\sum _{\mathrm{p}=\mathrm{0}}^{\mathrm{p}}\sum _{\mathrm{k}=\mathrm{0}}^{\mathrm{K}}\sum _{\mathrm{j}=\mathrm{0}}^{\mathrm{J}}{\mathrm{w}}_{\mathrm{p},\mathrm{k},\mathrm{j}}{\left|{\mathrm{S}}_{\mathrm{i}}\left(\mathrm{n}-\mathrm{k}\right)\right|}^{\mathrm{p}}{\mathrm{PS}}_{\mathrm{i}}\left(\mathrm{n}-\mathrm{j}\right)$$

  

  where S_i is a model input signal; j is signal delay of an input signal, k is signal delay of an input signal modulus value; p is a model order; J is maximum signal delay of the input signal, k is maximum signal delay of the input signal modulus value; P is the highest model order; and W_{p, k, j} is a model coefficient. Correspondingly, in the present invention, S_i is an output signal of a signal generator module, namely the main link signal; S_o is a target matrix, namely the feedback signal.
• In order to get a correction parameter W, a common calculation method called LS algorithm is generally used to obtain: $$\mathrm{W}={\mathrm{R}}^{-\mathrm{1}}\mathrm{Z}$$

  

  $$\mathrm{R}=\sum _{\mathrm{p}=\mathrm{0}}^{\mathrm{p}}\sum _{\mathrm{k}=\mathrm{0}}^{\mathrm{K}}\sum _{\mathrm{j}=\mathrm{0}}^{\mathrm{J}}{\left|{\mathrm{S}}_{\mathrm{i}}\left(\mathrm{n}-\mathrm{k}\right)\right|}^{\mathrm{p}}{\mathrm{S}}_{\mathrm{i}}\left(\mathrm{n}-\mathrm{j}\right)$$

  

  where R is a parameter extraction matrix; Z is a target matrix, namely the target matrix So in the formula (6); W is a correction parameter; (•)-1 is a pseudo-inverse operation; and meanings of other expressions are the same as in the formula (6).
• It should be noted that, methods for forming the parameter extraction matrix Rr and the target matrix Zr corresponding to the main link signal and the parameter extraction matrix Rs and the target matrix Zs corresponding to the robustness signal source are the same, and are all generated according to the formula (7) and formula (8).
• Here, the derivation formula (8) for R has been given, i.e., it is generated by combining two items in the memory multinomial model (6), and all the physical meanings in the formula are defined expressly.
• In the combination of the matrixes, the matrix composition unit performs matrix composition using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

  

  $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

  

  where, a and b are weighting coefficients.
• A correction parameter generation unit is arranged to extract a linearization parameter according to the matrixes R and Z and use the linearization parameter as the correction parameter to be loaded into the pre-corrector. Generally the linearization parameter is extracted using adaptive algorithms, such as a least squares (LS) algorithm, a recursive least squares (RLS) algorithm or a least mean square (LMS) algorithm and the like, the extracted linearization parameter is downloaded into the pre-corrector, and the correction parameter is updated and downloaded into a correction information generation unit.
• As shown in FIG. 4, the pre-corrector includes:
• an index address acquisition unit arranged to calculate an amplitude value or power of the main link signal to obtain an index address of a correction signal, linear or nonlinear mapping (bit truncating operation) being performed on the amplitude value or power of the input signal to generate address information, perform addressing and obtain the index address;
• a correction signal acquisition unit arranged to look up a corresponding lookup table (linearization lookup table) according to the index address and obtain the correction signal according to contents of the lookup table; and
• a pre-correction processing unit arranged to perform the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter, where the pre-correction processing unit performs pre-correction processing according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$
  
  $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$
  
  $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$
  
  where, y is a pre-corrected signal; n is a signal sampling time serial number; F_U,X(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_U,X(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
• The pre-corrector pre-corrects the main link signal according to amplitude and phase information of the signal, where the amplitude is equal to amplitude of a distortion signal generated by the nonlinear system and the phase is opposite to phase of the distortion signal, thus the distortion caused by the nonlinear system to the main link signal can be counteracted
  conversion of the pre-corrected signal from a digital domain to an analog domain is implemented through the DAC, and nonlinear processing of the signal is implemented through the nonlinear system. The output signal of the nonlinear system is ultimately turned into a fedback digital signal through the ADC. The emphasis of the technical scheme of the present invention is not the DAC, nonlinear system and ADC, thus the embodiments of the present invention can be accomplished by using the existing modules above. Therefore, the above devices will not be described in detail in the embodiments of the present invention.
• FIG. 5 is a specific embodiment of an apparatus for power amplifier pre-correction (an apparatus for nonlinear system distortion correction) including a signal generator module, a channel filtering module, a pre-corrector module, a DAC module, an ADC module, an up-conversion module, a down-conversion module, a local oscillator (LO) module, a power amplifier module, an attenuator module, an adaptive device module and a robustness signal source module. Compared to FIG. 2, in the implementation of the pre-corrector module, the pre-corrected signal is obtained according to the pre-correction function shown in the formula (6). In a correction parameter extraction unit of the adaptive device module, upon construction of a parameter extraction matrix, parameter extraction matrixes R obtained by the current link signal and the robustness signal source are both constructed according to the formula (8), and the matrixes R and Z eventually used for extracting the parameters are constructed according to the formulas (4) and (5).
• As shown in FIG. 6, an embodiment of the present invention further relates to a method for nonlinear system distortion correction implemented by the above apparatus including the following steps:
• Step 601: generating a robustness signal provided to an adaptive device;
• Step 602: collecting a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and
• Step 603: performing, according to the correction parameter, pre-correction processing on the main link signal output by the signal generator.
• Further, generating a correction parameter specifically includes:
• collecting the main link signal output by the signal generator and the feedback signal returned by the main link signal via the feedback link;
• generating a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generating a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal, and performing matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z; and
• extracting a linearization parameter according to the matrixes R and Z, and using the linearization parameter as the correction parameter to be loaded into a pre-corrector.
• The pre-correction processing specifically includes:
• calculating an amplitude value or power of the main link signal to obtain an index address of a correction signal;
• looking up a corresponding lookup table according to the index address, and generating the correction signal according to contents of the lookup table; and
• performing the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter; in the step, the pre-correction processing is performed according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$
  
  $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$
  
  $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$
  
  where, y is a pre-corrected signal; n is a signal sampling time serial number; F_U,X(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_U,X(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
• Matrix composition is performed using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

  

  $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

  

  where, a and b are weighting coefficients.
• Generating a robustness signal includes:
• performing frequency conversion, filtering and frequency shift processing on a baseband signal to obtain the main link signal and performing data property processing on the main link signal to obtain the robustness signal. Further, after collecting the main link signal and the feedback signal, the data property processing is performed on the main link signal and the feedback signal. The data property processing refers to performing operations of delay aligning and energy aligning on the collected signals.
• A digital signal after the pre-correction processing is converted into an analog signal.
• Nonlinear processing is performed on the analog signal.
• A signal is coupled out from the signal after the nonlinear processing and is converted into a digital signal to be returned as the feedback signal.
• The technical scheme of the present invention is not limited to the application in a TD-SCDMA and TDL-LTE mode mixing cross-band ultra-wideband RRU system. For a Global System for Mobile communication (GSM), Code Division Multiple Access (CDMA), Universal Mobile Telecommunications System (UMTS), Frequency Division Duplexing-Long Term Evolution (FDD-LTE) and Worldwide Interoperability for Microwave Access (WiMAX) single-mode or mixed-mode system, the effect of improving the robustness of the system using the power amplifier linearization technology is remarkable as well. The technical scheme of the present invention has no particular requirements on signal modulation, bandwidth and the like, and is commonly applied to improve the robustness of various systems using the power amplifier linearization technology, which is obviously superior to the robustness of traditional systems using the power amplifier linearization technology, greatly improving the working security and performance robustness of communication systems and even the entire wireless base station system.
• Although the preferred embodiments of the present invention have been disclosed for the purpose of illustration, those skilled in the art will realize that various improvements, additions and replacements are also possible. Therefore, the scope of the present invention should not be limited to the above embodiments.
Industrial Applicability
• As mentioned above, an apparatus and method for nonlinear system distortion correction provided by the embodiments of the present invention have the following beneficial effects: by providing a robustness signal, it is ensured that a system will not bring the problem of unknowable abrupt change in power or abnormal output frequency spectrum, which is caused by the fact that the difference between a signal characteristic in a former configuration and a signal characteristic in a latter configuration is too large under various dynamic configurations after an actual link signal passes through a pre-distortion module, thereby greatly improving the robustness of the system in the case of dynamic operation, and being able to satisfy various frequency band configuration requirements of operators and dynamic change requirements of telephone traffic in different regions.

Claims (18)

1. An apparatus for nonlinear system distortion correction comprising:

   a robustness signal source generator, arranged to generate a robustness signal provided to an adaptive device;

   the adaptive device, arranged to collect a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link, and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and

   a pre-corrector, arranged to according to the correction parameter perform pre-correction processing on the main link signal output by the signal generator.
2. The apparatus according to claim 1, wherein, the adaptive device comprises:

   a data collection unit, arranged to collect the main link signal output by the signal generator and the feedback signal returned by the main link signal via the feedback link;

   a matrix composition unit, arranged to generate a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generate a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal, and perform matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z; and

   a correction parameter generation unit, arranged to extract a linearization parameter according to the matrixes R and Z, and use the linearization parameter as the correction parameter to be loaded into the pre-corrector.
3. The apparatus according to claim 1 or 2, wherein, the pre-corrector comprises:

   an index address acquisition unit, arranged to calculate an amplitude value or power of the main link signal to obtain an index address of a correction signal;

   a correction signal acquisition unit, arranged to look up a corresponding lookup table according to the index address, and generate the correction signal according to contents of the lookup table; and

   a pre-correction processing unit, arranged to perform the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter.
4. The apparatus according to claim 3, wherein, the pre-correction processing unit performs the pre-correction processing according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$

   

   $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$

   

   $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$

   

   wherein, y is a pre-corrected signal; n is a signal sampling time serial number; F_U,X(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_U,X(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
5. The apparatus according to claim 2, wherein the matrix composition unit performs the matrix composition using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

   

   $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

   

   wherein, a and b are weighting coefficients.
6. The apparatus according to claim 1, wherein, the robustness signal source generator is arranged to generate a robustness signal by the following way:

   performing frequency conversion, filtering and frequency shift processing on a baseband signal to obtain the main link signal, and performing data property processing on the main link signal to obtain the robustness signal.
7. The apparatus according to claim 6, wherein, the adaptive device performs the data property processing on the main link signal and the feedback signal after collecting the main link signal and the feedback signal.
8. The apparatus according to claim 7, wherein, the data property processing refers to performing operations of delay aligning and energy aligning on the collected signals.
9. The apparatus according to claim 1, 2, 4, 5 or 8, wherein, the apparatus further comprises:

   a digital to analog converter, arranged to convert a digital signal output by the pre-corrector into an analog signal;

   a nonlinear system, arranged to perform nonlinear processing on the analog signal; and

   an analog to digital converter, arranged to couple a signal out from the output signal by the nonlinear system, and convert the signal into a digital signal which is used as the feedback signal to be input into the adaptive device.
10. A method for nonlinear system distortion correction comprising:

    generating a robustness signal provided to an adaptive device;

    collecting a main link signal output by a signal generator and a feedback signal returned by the main link signal via a feedback link, and generate a correction parameter according to the main link signal, the feedback signal and the robustness signal; and

    performing, according to the correction parameter, pre-correction processing on the main link signal output by the signal generator.
11. The method according to claim 10, wherein, generating a correction parameter further comprises:

    collecting the main link signal output by the signal generator and the feedback signal returned by the main link signal via the feedback link;

    generating a linearization parameter extraction matrix Rr and a target matrix Zr according to the main link signal and the feedback signal, generating a linearization parameter extraction matrix Rs and a target matrix Zs according to the robustness signal, and performing matrix composition on the Rs, Zs and the Rr, Zr to obtain matrixes R and Z; and

    extracting a linearization parameter according to the matrixes R and Z, and using the linearization parameter as the correction parameter to be loaded into a pre-corrector.
12. The method according to claim 10 or 11, wherein, the pre-correction processing further comprises:

    calculating an amplitude value or power of the main link signal to obtain an index address of a correction signal;

    looking up a corresponding lookup table according to the index address, and generating the correction signal according to contents of the lookup table; and

    performing the pre-correction processing on the main link signal output by the signal generator according to the correction signal and the correction parameter.
13. The method according to claim 12, wherein, the pre-correction processing is performed according to formulas (1), (2) and (3): $$y\left(n\right)=F_{U,\mathit{X}}\left(U,\mathit{X}\right)$$

    

    $$\mathit{U}=\left[\mathit{U}\left(\mathit{n}\right),\mathit{U}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{U}\left(\mathit{n}-\mathit{K}\right)\right]$$

    

    $$\mathit{X}=\left[\mathit{x}\left(\mathit{n}\right),\mathit{x}\left(\mathit{n}-\mathit{1}\right),\cdots ,\mathit{x}\left(\mathit{n}-\mathit{J}\right)\right]$$

    

    wherein, y is a pre-corrected signal; n is a signal sampling time serial number; F_U,X(•) is a pre-correction function, the correction parameter being a coefficient in the pre-correction function F_U,X(•); U is a vector of the correction signal; X is a vector of the main link signal; K is maximum delay of the correction signal; and J is maximum delay of the main link signal.
14. The method according to claim 11, wherein, the matrix composition is performed using formulas (4) and (5): $$R=\mathrm{a}\times \mathrm{Rs}+\mathrm{b}\times \mathrm{Rr}$$

    

    $$Z=\mathrm{a}\times \mathrm{Zs}+\mathrm{b}\times \mathrm{Zr}$$

    

    wherein, a and b are weighting coefficients.
15. The method according to claim 10, wherein, generating a robustness signal comprises:

    performing frequency conversion, filtering and frequency shift processing on a baseband signal to obtain the main link signal and performing data property processing on the main link signal to obtain the robustness signal.
16. The method according to claim 15, wherein, after collecting the main link signal and the feedback signal, the method further comprises:

    performing the data property processing on the main link signal and the feedback signal.
17. The method according to claim 16, wherein, the data property processing refers to performing operations of delay aligning and energy aligning on the collected signals.
18. The method according to claim 10, 11, 13, 14 or 17, wherein,
    a digital signal after the pre-correction processing is converted into an analog signal; nonlinear processing is performed on the analog signal; and
    a signal is coupled out from the signal after the nonlinear processing and is converted into a digital signal to be returned as the feedback signal.

Images